CHEMISTRY. 21 



On the detection of nitrous acid in water by means of amidonaplitholsul- 

 phonic acid according to Erdmann, II. ]Mexxicke [ZLscIu: Aiujfw. Chein., 1900, pp. 

 '235, '236; ahx. In ('hern. Centbl., 1900, I, No. 13, p. 733). — Tests are reported which 

 indicate that ]'>(lmann's method (see above) is a delicate and rehable means of 

 detectiuj,^ and determining nitrites in water. Some precautious to l>e observed in 

 manipulatinii ai'e ex])lained. 



The value of methods for detecting nitrites in drinking water, L. {Spiegel 

 {Ber. Deut. Cheni. Gesell., 33 {1900), pp. 639-644; aba. in Cheiii. Ztg., 24 {1900), No. 32, 

 liepevt. , p. 113; Jour. Chem. Soc \_London'\ , 78 { 1900) , No. 450, II, p. 318) .—The author 

 tested Erdmaim's method (see above) in comparison with various other methfxls 

 for the same i)urpose and found it less sensitive than the potassium iodid starch 

 inetliod (If tlic Lunge-Elosvay reagent. The author considers determinations of 

 nitrites in drinking water as of little value, since the nitrites represent a transition 

 stage and their determination may be misleading as to normal conditions. Their 

 detection, however, may be of a negative value, and for this purpose the author rec- 

 ommenils the use of guaiacol or creosote, which, in presence of nitrites in dilute solu- 

 tion, give an orange and a yellow coloration respectively. These reactions are not 

 interfered with by the presence of oxidizing agents, such as nitrates, chlorates, and 

 hydrogen peroxid, or Ijy ferric salts in amounts usually met with inpotable. waters. 



On a simple method for determining phosphoric acid in connection w^ith 

 metabolism experiments, A. Neumanx {Ardi. Anat. v. Plu/slol., Phiisial. Ahl., 1900, 

 l>/>. 159-105; (lbs. in Chem. Centbl., 1900, I, No. 10, p. 571). 



Estimating the w^ater in cereals — practical methods, J. F. Hoffmann 

 {]Vclin.'<rhr. Bnin., 16 (1899), jip. 569-574, 585-588, 605). 



The determination of the sugar content of molasses feeding stuffs, A. >1en- 

 ZEi. {Di'iit. Znckeriad., 25 {1900), No. 14, pp. 552, 553). 



The preparation of a nonsugar from beet juice, A. Ri'mpleh {Dent. Znckerind., 

 25 {1900), No. 15, pp. 592, 593). - 



A comparison of some formaldehyde tests, B. M. Pilhashv {Jonr. Ann'r. ('hem. 

 Soc, 22 {1900), No. 3, pp. 132-135). — Several tests for small quantities were com- 

 pared. Phenylhydrazin hydrochlorid seemed to be the best reagent. A solution of 

 1 gm. of phenylhydrazin hydrochlorid with 1.5 gm. sodium acetate in 10 cc. of water 

 was used. To 1 cc. of the liquid to be tested 2 drops of the reagent and 2 dro])S of 

 sulphuric acid are added, giving a green coloration if formaldehyde is present. In 

 weak solutions (1: 10,000 to 1 : 250,000) take 3 cc. of the liquid and 4 or 5 drops of the 

 reagent and of sulphuric acid, heating if necessary. 



Tests for the strength of solutions of formaldehyde, H. A. Huston {Indiana 

 Sia. Rpt. 1899, pp. 76, 77). — The formaldehyde content of 5 samples of commercial 

 formalin was determined from the specific gravity and ])y the ammonia and the 

 potassium cyanid methods. The results are tabulated. Brief notes are given on 

 different methods of analysis. The table in Allen's Commercial Organic Analysis 

 used in calculating the formaldehyde content from the specific gravity was not con- 

 sidered appliral)le to the sanq)les examined. 



On Hubl's iodin method for oil analysis, A. H. Gili, and "\V. (). Adams {.Tone. 

 Amer. Cliem. So<:, 22 {1900), No. 1, ]>p. 12-14). 



A new^ method for the determination of fat in dairy products, Lixdet { lud. 

 Lait., 25 {1900), No. 23, pp. 177, 178, Jig. 1). — The method depends upon the solubilit}' 

 of casein in a concentrated solution of resorcin. The sample of milk or cheese is treated 

 with a strong solution of resorcin with the addition of a few drops of alkali. This is 

 heated in a water bath until the separation of the fat is completed. The apparatus 

 devised for the purpose consists in part of a graduated cylinder from which the 

 reading for fat is taken. The addititm of some coloring matter for clearly differ- 

 entiating the fat layer is recommended. 



